Публикация:Корогодин 2019 Triangulation Positioning by Means of Wi-Fi Signals in Indoor Conditions
Ilya V. Korogodin, Vladimir V. Dneprov, Olga K. Mikhaylova Triangulation Positioning by Means of Wi-Fi Signals in Indoor Conditions // 2019 Progress in Electromagnetics Research Symposium (PIERS). — 2019.
| BibTeX: |
@article{korogodinPIERS2019SWiFi,
author = "Ilya V. Korogodin and Vladimir V. Dneprov and Olga K. Mikhaylova",
title = "Triangulation Positioning by Means of Wi-Fi Signals in Indoor Conditions",
journal = "2019 Progress in Electromagnetics Research Symposium (PIERS)",
doi = "10.1109/PIERS-Spring46901.2019.9017863",
year = "2019",
language = english
}
|
Аннотация
A convenient indoor navigation is still an unsolved problem. The solution is in high demand: advertising, goods promotion, subway navigation and so on need a cheap and reliable positioning method. WiFi-based positioning looks like a promising candidate: it doesn’t require an additional complex local infrastructure, user devices (mobile phones) can process Wi-Fi signals right now, Wi-Fi signals are wideband and strong. Wi-Fi positioning methods can be divided into categories: fingerprinting, received signal strength indication, time-of-flight and triangulation based. In the study we considered a triangulation method. Several methods to form signal angle-of-arrival measurements are known: MUSIC, generalized cross-correlation and others. In addition, in the study we presented a maximum likehood angle estimation algorithm for OFDM Wi-Fi signals and multi-antenna receivers. The algorithms have differences. Proposed one allows to measure angle-of-arrival and anlge-of-departure, it has low estimation noise. The MUSIC algorithm can determine multipath propagation. The generalized algorithm is simple enough. But all the algorithms utilize the antenna mathematical model. The model provides the relationship between signal phase differences for antennas and the angle- of-arrival. The algorithms need the appropriate model to get good positioning accuracy. We compared two models: a model of independent zero-size antennas and a finite-size model of mutual coupled antennas. The first one is simple enough and is described by trigonometric formulas. The second one is complex, and we used electromagnetic simulation into CST Studio to get phase-to-angle relations. The electromagnetic simulation predicts up 5-10 degree corrections to angle-of-arrival in comparison with simple independent antennas model. To approve simulation results we created Wi-Fi receiver and transmitter prototypes based on COTS Intel 5300 modules and custom drivers CSITool. It is shown by the experiments that the antennas mutual coupling offsets the angle of arrival estimations about 5-10 degrees. We performed the experiments in our laboratory room and got from 3 to 10 degrees AoA estimation error profile. The error profile was used for a triangulation positioning simulation, as result we achieved about 1 m localization error.
Галерея
Принт статьи - PDF.
Презентация - PDF.
